Real Time Filtering using DSP56311EVM

I have a basic question:

I was performing real-time filtering of a music wav file on the
DSP56311EVM using Signal Wizard.My  low pass filter had 1015 taps, sampled
at 48Khz, single channel. 

The filter worked fine.

However when i used the DSP Software to change the filter to Dual Channel
Mode, the filter failed to work. 

I thought the DSP could handle 300 million multiply accumulates per
second. 
I worked out mine to be (48Khz x 1015 x 2 Channels  = 97 million MACs)

Anyone one know why it didn't work ?

Thanks

ndegwakg wrote:
> I have a basic question:
> 
> I was performing real-time filtering of a music wav file on the
> DSP56311EVM using Signal Wizard.My  low pass filter had 1015 taps, sampled
> at 48Khz, single channel. 
> 
> The filter worked fine.
> 
> However when i used the DSP Software to change the filter to Dual Channel
> Mode, the filter failed to work. 

"failed" is such a loose term ;)

How did it fail:
   1. caused your house to burn down?
   2. no output?
   3. garbled output?
   4. noisy output?
   etc etc etc

What diagnostic attempts have you made:
   1. zero input gives what result?
   2. pure tone on only one channel results in what output?

I've *NO* familiarity with the DSP56311EVM, I'm trying to give you an 
idea of the types of questions you should be asking yourself.

The basic approach to solving a problem is "Divide and Conquer", a 
corollary to Newton's Method if you wish.

A side issue:
What is your background? That would assist those wishing to assist to 
respond in your universe of discourse.

HTH




> 
> I thought the DSP could handle 300 million multiply accumulates per
> second. 
> I worked out mine to be (48Khz x 1015 x 2 Channels  = 97 million MACs)
> 
> Anyone one know why it didn't work ?
> 
> Thanks
> 
> 
> 
> 

>ndegwakg wrote:
>> I have a basic question:
>> 
>> I was performing real-time filtering of a music wav file on the
>> DSP56311EVM using Signal Wizard.My  low pass filter had 1015 taps,
sampled
>> at 48Khz, single channel. 
>> 
>> The filter worked fine.
>> 
>> However when i used the DSP Software to change the filter to Dual
Channel
>> Mode, the filter failed to work. 
>
>"failed" is such a loose term ;)
>
>How did it fail:
>   1. caused your house to burn down?
>   2. no output?
>   3. garbled output?
>   4. noisy output?
>   etc etc etc
>
>What diagnostic attempts have you made:
>   1. zero input gives what result?
>   2. pure tone on only one channel results in what output?
>
>I've *NO* familiarity with the DSP56311EVM, I'm trying to give you an 
>idea of the types of questions you should be asking yourself.
>
>The basic approach to solving a problem is "Divide and Conquer", a 
>corollary to Newton's Method if you wish.
>
>A side issue:
>What is your background? That would assist those wishing to assist to 
>respond in your universe of discourse.
>
>HTH
>
>
>
>
>> 
>> I thought the DSP could handle 300 million multiply accumulates per
>> second. 
>> I worked out mine to be (48Khz x 1015 x 2 Channels  = 97 million MACs)
>> 
>> Anyone one know why it didn't work ?
>> 
>> Thanks
>> 
>> 
>> 
Oops sorry, i'll try to be clearer.

The DSP56311EVM (a Motorola DSP Hardware Module) together with the
software 'Signal Wizard' are supposed to perform real-time filtering. The
software allows you to design the filter specs e.g no of taps, sample rate,
type of windowing function etc. on the PC 

The designed filter is then downloaded on to the hardware (DSP56311EVM)
via a serial cable connecting the PC and the DSP Hardware. A *.wav file
from the PC is then passed to the hardware and is filtered in real time
giving an output at the DSPs line out port.

I tried doing 

Diagnostic Attempts Done:
1. When the filter is designed to use 511 taps, dual channel, it works as
expected i.e the output at the DSP Line Out is the filtered version of the
input wav file
2. When the filter is designed to use 1015 taps, dual channel, it fails
i.e the output at the DSP Line Out is not the filtered version. It passes
all the frequency components of my input as if it were an all-pass filter
3.When the filter is designed to use 1015 taps, single channel, it WORKS.

So at least i've succeeded in finding out that the number of taps plus the
number of Channels are affecting how the DSP chip is processing the
filter.

What i don't know is how and why ?

I hope i'm clearer now.


(Aside: I took a DSP course during my Undergrad 2 years ago. )